Purpose: Results from different costing studies vary greatly because they rely on different modeling techniques, secondary data sources or on numerous assumptions to project potential long-term costs associated with the treatment of diabetes. The objective of this study is to estimate the short-term and long-term healthcare costs of diabetes and diabetes-related complications for a large cohort of previously diagnosed and newly diagnosed patients followed prospectively.

Methods: All incident and prevalent cases of diabetes in Ontario from 1992-2002 were followed for up to 10 years, until death, or out migration. This Ontario Diabetes Database was linked to various healthcare administrative datasets to develop comprehensive costing of diabetes and diabetes-related complications. Seven diabetes complications were tracked over time (i.e. ischemic heart disease, myocardial infarction, heart failure, stroke, amputation, renal failure, and blindness). Costs were compiled for outpatient services, pharmaceuticals, long-term care, home care and hospitalizations for each year of follow-up. Regression models were used to estimate the probability of incurring some healthcare costs and to estimate the total healthcare costs conditional on at least one event occurring.

Results: There were 498,590 incident cases and 734,113 total patients with diabetes in Ontario in the database and over 1.39 million diabetes-related events, of which 95% were cardiovascular in nature. Of the newly diagnosed patients without complications, the mean healthcare cost was $3,115 in the year of diagnosis and approximately $2,109 in subsequent years. On average, complications resulted in an additional $2,062 per patient per year of follow-up. Generally, the cost of a complication in the year of the event was substantially higher than in subsequent years (e.g. cost of amputation in the first year was $34,469 and $4,721 in following years). The estimated costs differ as the demographic and clinical characteristics of individual patients vary. For incident cases, the annual probability of diabetes-related events remained fairly constant over the study period and the 10-year cumulative probability of death was 47%.

Conclusions: The results confirm the high cost of seven diabetes-related complications. Our large sample and ability to link administrative databases provide a unique opportunity to estimate the incidence of events and the high cost of diabetes over time.